Scaling Intercompartmental Clearance and Flow

Why Scale Intercompartmental Clearance and Flow Carefully?

Intercompartmental clearance determines how rapidly drug is exchanged between compartments in a pharmacokinetic model. Physiological flows, such as blood flow through an organ, describe volumetric transport through anatomy. Both can have units of volume per time, but that shared unit does not make them the same biological quantity or guarantee that they have the same relationship to body size.

The practical question is therefore not simply whether a parameter can be scaled. It is what the parameter represents, what mechanism might connect it to body size, and whether the available data support that scaling assumption. The earlier material on allometric scaling and on why clearance and volume need different scaling provides the general framework; here the emphasis is on exchange clearance and physiological flow.

Reference-Normalized Scaling Models

Let \(W_i\) be individual \(i\)'s body-size measure and \(W_{\mathrm{ref}}\) a stated reference size. A basic allometric model for intercompartmental clearance is

Here \(Q_{\mathrm{ref}}\) is the typical clearance at the reference size, and \(\theta_Q\) is the size exponent. The ratio \(W_i/W_{\mathrm{ref}}\) is dimensionless, so \(Q_i\) retains the units of \(Q_{\mathrm{ref}}\). At \(W_i=W_{\mathrm{ref}}\), the ratio is one and the model gives \(Q_i=Q_{\mathrm{ref}}\), whatever the exponent.

If the NLME model includes between-individual variation on \(Q\), one possible positive-parameter specification is

This adds an individual random effect on the log scale. It does not change the units of \(Q_i\), and it preserves positivity. The size relationship describes the typical size trend; the random effect represents individual deviations from that trend. Whether to include such a random effect, and its relationship to other random effects, is a modeling decision rather than a consequence of allometry.

The same mathematical form can be used for a physiological flow:

The exponent \(\theta_{F_j}\) need not equal \(\theta_Q\). A physiological flow may be informed by organ perfusion or whole-body circulatory physiology, while \(Q\) is an exchange parameter in the chosen pharmacokinetic model. The value of either exponent should be justified for the specific parameter and population. A familiar exponent is not automatically appropriate merely because the parameter has units of volume per time.

How Exchange Clearance Enters the Compartment Equations

The meaning of \(Q\) becomes clearer by writing the exchange terms. Let \(A_c\) and \(A_p\) denote drug amounts in the central and peripheral compartments, and let \(V_c\) and \(V_p\) be their volumes. The corresponding concentrations are \(C_c=A_c/V_c\) and \(C_p=A_p/V_p\). For exchange alone, the amount rates are

The same \(Q\) appears in both directions. The terms \(QC_c\) and \(QC_p\) have units of amount per time, as required for derivatives of amounts. Adding the equations gives

Thus exchange transfers amount between compartments but does not create or destroy total drug amount. Rewriting each equation in terms of amounts gives the exchange coefficients \(Q/V_c\) and \(Q/V_p\), each with units of inverse time. These are not separate clearances: both arise from the one shared exchange clearance and the corresponding compartment volumes.

Scaling Physiological Flow Without Renaming It

A flow parameter should be identified by what is physically flowing, where it flows, and how it enters the model. For example, a model might use \(F_j\) for blood flow through a specified organ. If the reference flow is measured or otherwise justified at \(W_{\mathrm{ref}}\), a size model for that flow can be written using \(F_{j,\mathrm{ref}}\) and \(\theta_{F_j}\). The units alone cannot tell whether that \(F_j\) is an organ blood flow, an intercompartmental clearance, or another model parameter.

A flow can enter a model as a transport rate, while a clearance parameter represents a volume-per-time relationship between concentration and amount transfer or elimination. In a compartment model, \(Q\) is defined by the exchange terms, not by an anatomical claim that a measured blood stream of volume \(Q\) passes between two compartments. A physiological flow may help motivate a model parameter in a specified mechanistic structure, but equating the two requires a model-based argument.

Choosing and Checking the Assumption

A scaling exponent is a scientific assumption, whether fixed from prior knowledge or estimated from data. Before using one, check that the reference parameter and target parameter have the same definition and units, that \(W_{\mathrm{ref}}\) and the body-size measure are clearly specified, and that the model is intended for the range of sizes being analyzed. An estimated exponent can be weakly informed if the observed sizes cover only a narrow range or if size covaries strongly with other relevant characteristics.

Diagnostics should also be tied to the assumed relationship. Useful checks include plotting individual or typical parameter estimates against body size, inspecting residual patterns after scaling, and checking whether the fitted size trend is dominated by a few subjects at the extremes. Such plots do not by themselves prove a causal physiological relationship, and estimated individual parameters may be noisy. A systematic remaining trend can indicate a poor exponent, an unsuitable size descriptor, an omitted covariate, or a structural model mismatch; it does not identify which explanation is correct without further investigation.

Extrapolation deserves particular caution. A model that behaves plausibly across the observed adult weight range may be unsupported at much smaller sizes, where anatomy and physiology can differ beyond simple proportional or allometric size effects. The body-size model also should not be used to conceal a mismatch in parameter meaning—for example, treating an exchange clearance as measured organ perfusion solely because their units agree.

Common Pitfall and Model Alternatives

A common mistake is to reason, “\(Q\) and organ blood flow are both measured in L/h, so they should have the same exponent and can be substituted for one another.” The correction is that units establish dimensional compatibility, not biological identity. An exponent for \(Q\) must be justified as an assumption about the exchange parameter; a flow exponent must be justified for the particular physiological flow. The two may coincide in a particular model, but the units do not establish that result.

The main alternatives are to fix an exponent based on a prespecified mechanistic or empirical rationale, estimate an exponent from sufficiently informative data, or omit size scaling when there is no defensible or identifiable relationship. A fixed exponent can stabilize estimation but may impose a wrong trend. Estimating an exponent allows data to inform the relationship but can be imprecise, especially over a limited size range. Omitting scaling avoids an unsupported assumption but may leave systematic size-related variation unexplained. These choices should be compared in terms of parameter meaning, data support, and model adequacy—not selected solely by which gives a preferred numerical estimate.